Key Takeaways:
This interactive table presents ICU patient variables with clear labels
and units for easy interpretation. Users can explore, sort, and filter
key clinical variables to better understand patient characteristics.
Key Takeaways:
This plot compares ICU admission time across age groups and highlights gender differences using color. Each dot represents an individual patient, revealing the distribution and variation within each group. Older age groups show greater variability in admission times, while gender differences appear minimal.
Key Takeaways:
This plot compares ICU admission probability over time between sepsis and non-sepsis patients, using color to distinguish the two groups. The interactive tooltips allow users to explore how admission probability changes by time and sepsis status, providing a clearer view of group differences. ICU admission probability drops more steeply for sepsis patients in the early hours after hospital admission.
These visualizations are based on the ICU patients dataset from The PhysioNet/Computing in Cardiology Challenge 2019, which provides data on 40,336 sepsis and non-sepsis patients. The dataset was collected from the electronic medical record systems of Beth Israel Deaconess Medical Center and Emory University Hospital over the past decade. After removing observations with missing key variables, a total of 34,750 records were retained.
The link to the GitHub repository is https://github.com/Menglong-Yang/flexdashboard-icu.
This dashboard helps doctors understand how different vital signs and lab results are related to ICU admissions. The findings is valuable for healthcare providers to better identify high-risk patients who need ICU care urgently.